1. Python based particle tracking code for monitor design 20160303 Kenichirou Satou J-PARC/KEK

2. Tokai, Ibaraki 30 GeV MR 340 kW (750 kW) MLF Hadron hall detector 3 GeV RCS 600 kW (1 MW) LINAC 30 mA (operational) (50 mA (goal)) J-PARC Japan Proton Accelerator Research Complex

3. LINAC Rapid Cycling Synchrotron (RCS) Main ring (MR) MLF Neutrino Hadron http://j-parc.jp/Acc/ja/about.html Ion Source Accelerators in J-PARC

4. Main ring (MR) 500 m Injection Fast extraction (FX) Slow extraction (SX) RF cavity Neutrino Hadron Beam orbit tuning with accurate beam profile information High intensity ⇒ reduction of beam loss IPM (H/V) IPM (H) Total length (m)1568 Energy (GeV)330  0.97120.9995 Lorentz  4.2032.97 Harmonic number9 No. of bunches8 Circulating period (  sec) 5.385.23 RF frequency (MHz)1.671.72 Bunch length (time) (nsec) 20070 Bunch length (space) (m) 6020 Tune FX: x =22.40, y =20.75 SX: x =22.30, y =20.78

5. IPM (Horizontal) IPM (Vertical) Beam Injection Neutrino Hadron IPM (H/V) 540 m downstream from injection point Dispersion = 0

6. contents Position displacement? – Main Source? – Equation of motion of electron in IPM New python based simulation code Demonstration of the code No summary

7. Tracking Error Sources The error sources which induce position x displacement are,,,, Error E field: Space charge of the beam and guide E field – Ex=Esx+Egx – Ez=Esz(negligible small when relativistic beam)+Egz – Alignment error of the chamber Error B field: Magnet – Bx and Bz – Alignment error of the magnet Initial velocity of the detached electron (or ion) Electron-gas molecule collision -> Outside of this presentation Which one is the most important??? How much we should reduce the error source??? A simulation code will answer but it is worth while to check analytically

8. Eq. of motions (preliminary) And using Using the imaginary variables and parameters,,,,

9. Eq. of motions (preliminary) If α and θ is constant value On the detector plane Flight length Initial velocity Initial position is (0,0) Here,

10. Eq. of motions (preliminary) The position displacement due to initial TKE and error fields can be expressed as,

11. X position displacement due to error E field (preliminary) In case the magnetic field error is negligible small, the position x displacement can be expressed as, In a real IPM chamber, Gyro motion Initial TKE Gyro motion Field error: Ex, Ez E×B drfift (Ez×By) Field error: Ez

12. X position displacement due to error B field (preliminary) Geometrical shift Bx/B Gyro motion Field error: Bx/B, Bz/B E×B drfift (Ey×Bz) Field error: Bz B E

13. Python based particle tracking code Python based particle tracking code to design beam monitors which utilizes charged particles like IPM (detached electron and ions), Laser wire scanner (detached electron), Faraday cup based CT (secondary electron escape), etc. The fields in which the particles are moving, – Beam space charge effect Only 2D electric field: Relativistic beam condition – 3D electric and magnetic fields from another code which will be used to control the particle motion – Space charge effect of the tracked particles ensemble are not included -> single particle motion Was used to check CPS IPM (under development) performance

14. Python based 3D particle tracking code Profile simulator for IPM design Gaussian profile Successive Over Relaxation (SOR) to estimate Poisson eq. ・ Assumed to be 2D: Relativistic ・ Rectangular grid Grid data from POISSON/Superfish (2D) CST STUDIO SUITE (3D) ・ Rectangular or Cubic grid Ionization cross section Single differential cross section for H, He, H 2, CH 4, NH 3, and H 2 O Double differential cross section for H, and He Flow chart of the simulator 3D particle tracking 4th order Runge- Kutta method Beam intensity of bunch train

15. But in a real case, the profile could be far from the Gaussian-shape: J-PARC MR Simulation (100 kV, 70 kV) Space charge dilution by the 2 nd RF harmonics Transversal profile Longitudinal profile OTR@350BT WCM

16. SOR method Beam Potential Convergence at the Grids Default value About 470 iterations are needed to obtain a potential value at each grid point

17. Es calc.: Grid size selection Beam size: σx=σy=1mm

18. Tracking error vs. time step 4 th dorder Runge-Kutta

19. Initial conditions of a detached electron Ref. – "Differential cross section for ionization of helium,neon,and argon by high-velocity ions" J.H.Miller et. al., Phys. rev. A, 27 (1337), 1983. – "Analytic representation of secondary-electron spectra" J. Chem. Phys. 87 (12), 15 Dec. 1987 – "Electron production in proton collisions with atoms and molecules: energy distribution" M.E. Rudd et. al., Rev. Mod. Phys., 64, No2, 441 (1992) Single differential cross section (H, He, H2, CH4, NH3, H2O) Or Double differential cross section (H, He) are used for initial TKE of detached electron

20. An example of tracking Esx×By drift Gyro-motion ・ Initial momentum ・ Esx Initial position Final position

21. Cross check with PyECLOUD Only includes Esx,Esy Ecy By TKE=0 eV (Cross check with ESS code are now on going)

22. Calculation time using my 5 years old laptop PC SOR: Space charge E field -> 1079 s (18min) – Grid size: 70×70 – Conversion criteria: 1E-6 Tracking -> 124s (2min) / 1000 macroparticles – Time step: 0.005 ns – 600step/3ns ・ Intel Core i5 2.5GHz(4 CPUs): 64bit ・ 8.2GB memory TOFProfile

23. Demonstration How to use the code

24. In the near future upgrade Es calcu. – Non-Gaussian beam profile: SOR method – Analytical solution for Gaussian beam (Same as ESS code) – Non relativistic 3D Gaussian beam: Taking into account, Esz, Bs(x,y,z) fields